impact of arterial access route on bleeding complications

Circulation Journal Vol.79, September 2015

1928 SAITO S et al.Circulation JournalOfficial Journal of the Japanese Circulation Societyhttp://www.j-circ.or.jp

risk of readmission to treat recurrent bleeding, major adverse cardiovascular events (MACE), and all-cause mortality com-pared with patients without periprocedural bleeding.1 Most episodes of periprocedural bleeding are related to the vascular access route used.

Several trials in Europe/North America have compared the incidence of bleeding during PCI between the 2 main access

ercutaneous coronary intervention (PCI) is increasingly being used to treat acute coronary syndrome (ACS), resulting in a decrease in mortality and morbidity.

With the increased use of PCI and more intense anti-throm-botic regimens, periprocedural major bleeding has been rec-ognized as an important prognostic factor for cardiovascular mortality. Patients with periprocedural bleeding are at increased

P

Received March 10, 2015; revised manuscript received April 17, 2015; accepted April 26, 2015; released online June 4, 2015 Time for primary review: 19 days

Division of Cardiology, Shonan Kamakura General Hospital, Kamakura (S.S.); Ageo Central General Hospital, Ageo (T.I.); Graduate School of Medicine, Kyoto University, Kyoto (T.K.); Department of Cardiovascular Medicine, Graduate School of Medical Sciences, Kumamoto University, Kumamoto (H.O.); National Cerebral and Cardiovascular Center, Suita (H.O.); Fukuoka Sanno Hospital, Fukuoka (H.Y.); International University of Health and Welfare, Ohtawara (H.Y.); Clinical Research Support Center, Mie University Hospital, Tsu (M. Nishikawa); Division of Cardiology, Department of Medicine, Faculty of Medicine, Kinki University, Osakasayama (S.M.); Clinical Data and Biostatistics Department, R&D Division, Daiichi Sankyo, Co Ltd, Tokyo (Y.T.); and Division of Cardiovascular Medicine, Ohashi Medical Center, Toho University, Tokyo (M. Nakamura), Japan

PRASFIT-ACS: PRASugrel compared with clopidogrel For Japanese patIenTs with ACS undergoing PCI PRASFIT-Elective: PRASugrel compared with clopidogrel For Japanese patIenTs with coronary artery diseases undergoing Elective PCI

Mailing address: Shigeru Saito, MD, Division of Cardiology, Shonan Kamakura General Hospital, 1370-1 Okamoto, Kamakura 247-8533, Japan. E-mail: [email protected]

ISSN-1346-9843 doi: 10.1253/circj.CJ-15-0276All rights are reserved to the Japanese Circulation Society. For permissions, please e-mail: [email protected]

Impact of Arterial Access Route on Bleeding Complications in Japanese Patients Undergoing

Percutaneous Coronary Intervention– Insight From the PRASFIT Trial –

Shigeru Saito, MD; Takaaki Isshiki, MD; Takeshi Kimura, MD; Hisao Ogawa, MD; Hiroyoshi Yokoi, MD; Masakatsu Nishikawa, MD; Shunichi Miyazaki, MD; Yuko Tanaka;

Masato Nakamura, MD on behalf of the PRASFIT-ACS and PRASFIT-Elective Investigators

Background: Few large-scale studies have examined the relationship between bleeding events not related to coronary artery bypass grafting (CABG), and the vascular access route used in acute coronary syndrome (ACS) or in elective treatment of coronary artery disease (CAD).

Methods and Results: We compared the incidence of bleeding events occurring up to 3 days after percutaneous coronary intervention (PCI) or loading dose of prasugrel or clopidogrel in 2 studies of Japanese patients (PRASFIT-ACS, femoral and radial routes, n=683 and 531; PRASFIT-Elective, femoral and radial routes, n=135 and 508). Rates of periprocedural bleeding, bleeding not related to CABG, and puncture site bleeding were consistently lower in the radial access route group than in the femoral access route group in both studies. Risk factors for periprocedural bleeding included sex, body weight, age, and access route in PRASFIT-ACS (femoral access: hazard ratio [HR], 3.739; 95% confidence interval [CI]: 1.727–8.094; radial access: HR, 0.288; 95% CI: 0.128–0.65), and body weight, age, and access route in PRASFIT-Elective (femoral access: HR, 12.32; 95% CI 1.282–>100; radial access: HR, 0.125; 95% CI: 0.013–1.205).

Conclusions: The incidence of periprocedural bleeding is lower with a radial access route than with a femoral access route for PCI in Japanese patients with ACS or those undergoing elective PCI for CAD. (Circ J 2015; 79: 1928 – 1937)

Key Words: Arterial access route; Bleeding; Femoral artery; Percutaneous coronary intervention; Radial artery

ORIGINAL ARTICLECardiovascular Intervention


1929Arterial Access Route and Periprocedural Bleeding

Table 1. Baseline PRASFIT-ACS Patient Characteristics vs. Artery Access Route

Variable Femoral Radial Brachial Multiple* P-value

n 683 531 35 39

Female 147 (21.5) 103 (19.4) 6 (17.1) 9 (23.1) 0.364†

Mean age (years) 65.5±11.2 65.3±11.4 64.4±10.8 67.1±13.4 0.781†

Age ≥75 years 157 (23.0) 124 (23.4) 8 (22.9) 13 (33.3) 0.881‡

Height (cm) 162.4±8.7　　　　 163.0±8.9　　　　 161.7±7.3　　　　 161.4±9.7　　　　 0.283†

BW (kg) 63.3±11.3 65.4±12.0 65.0±10.1 66.4±14.6 0.002†

BW ≤50 kg 82 (12.0) 54 (10.2) 3 (8.6) 4 (10.3) 0.310‡

BW ≤60 kg 271 (39.7) 175 (33.0) 11 (31.4) 18 (46.2) 0.015‡

BMI (kg/m2) 23.9±3.3　　 24.6±3.6　　 24.8±3.1　　 25.3±4.0　　 0.001†

Disease type

UAP or NSTEMI 272 (39.8) 390 (57.1) 26 (74.3) 19 (48.7) <0.001‡　 STEMI 409 (59.9) 293 (42.9) 9 (25.7) 20 (51.3)

TIMI risk score

≤2 135 (19.8) 161 (23.5) 9 (25.7) 5 (12.8) 0.112‡

≥3 548 (80.2) 522 (76.5) 26 (74.3) 34 (87.2)

Duration of symptoms (min)

10–20 110 (16.1) 97 (18.3) 12 (34.3) 10 (25.6) 0.320‡

≥20 573 (83.9) 434 (81.7) 23 (65.7) 29 (74.4)

Time from onset of symptoms to PCI/LD (h)

≤3 178 (26.1) 88 (16.6) 8 (22.9) 7 (17.9) <0.001‡　 3–6 181 (26.5) 103 (19.4) 4 (11.4) 12 (30.8)

6–12 128 (18.7) 106 (20.0) 7 (20.0) 9 (23.1)

12–24 93 (13.6) 107 (20.2) 2 (5.7) 3 (7.7)

24–48 72 (10.5) 89 (16.8) 10 (28.6) 6 (15.4)

48–72 31 (4.5) 38 (7.2) 4 (11.4) 2 (5.1)

CrCl (ml/min)

>60 497 (72.8) 383 (72.1) 26 (74.3) 26 (66.7) 0.877‡

30–60 122 (17.9) 91 (17.1) 7 (20.0) 11 (28.2)

<30 7 (1.0) 4 (0.8) 0 1 (2.6)

Missing 57 (8.3) 53 (10.0) 2 (5.7) 1 (2.6)

Multiple treated lesions at initial PCI 188 (27.5) 144 (27.1) 9 (25.7) 22 (56.4) 0.875‡

Complications

Diabetes 227 (33.2) 207 (39.0) 19 (54.3) 16 (41.0) 0.038‡

Hyperlipidemia 497 (72.8) 419 (78.9) 27 (77.1) 32 (82.1) 0.014‡

Hypertension 485 (71.0) 397 (74.8) 27 (77.1) 34 (87.2) 0.145‡

Concomitant drugs used at the time of starting treatment

PPI 302 (44.2) 212 (39.9) 19 (54.3) 17 (43.6) 0.133‡

Ca antagonist 146 (21.4) 140 (26.4) 14 (40.0) 12 (30.8) 0.042‡

Statin 323 (47.3) 281 (52.9) 25 (71.4) 18 (46.2) 0.052‡

ACE inhibitor 104 (15.2) 84 (15.8) 4 (11.4) 8 (20.5) 0.777‡

ARB 167 (24.5) 147 (27.7) 13 (37.1) 13 (33.3) 0.202‡

β-blocker 108 (15.8) 108 (20.3) 5 (14.3) 5 (12.8) 0.041‡

Aspirin dose at time of LD (mg)

81–162 284 (41.6) 237 (44.6) 20 (57.1) 13 (33.3) 0.289‡

>162 399 (58.4) 293 (55.2) 15 (42.9) 26 (66.7)

CYP2C19 phenotype n=385 n=341 n=26 n=19

EM 140 (37.7) 120 (37.0) 12 (46.2) 4 (21.1) 0.039‡

IM 160 (41.6) 151 (44.3) 10 (38.5) 10 (52.6)

PM 80 (20.8) 64 (18.8) 4 (15.4) 5 (26.3)

Data given as mean ± SD or n (%). †t-test (femoral vs. radial); ‡χ2 test (femoral vs. radial). *Multiple access routes: femoral+radial, n=29; femoral+brachial, n=6; radial+brachial, n=3 femoral+radial+brachial, n=1. ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker; BMI, body mass index; BW, body weight; CrCl, creatinine clearance (estimated using the Cockcroft-Gault formula); EM, extensive metabolizer; IM, intermediate metabolizer; LD, loading dose; NSTEMI, non-ST-segment elevation myocardial infarction; PCI, percutaneous coronary intervention; PM, poor metabolizer; PPI, proton pump inhibitor; STEMI, ST-segment elevation myocardial infarction; TIMI, Thrombol-ysis in Myocardial Infarction; UAP, unstable angina pectoris.


1930 SAITO S et al.

with ACS undergoing PCI (PRASFIT-ACS)10 and PRASugrel For Japanese patIenTs with coronary artery diseases undergo-ing Elective PCI (PRASFIT-Elective)11 studies and then eval-uated the incidence of periprocedural bleeding in Japanese patients. These studies were performed to compare the inci-dence of bleeding in patients treated with adjusted doses of prasugrel (loading/maintenance dose [LD/MD], 20/3.75 mg) or clopidogrel (300/75 mg) following ACS or elective PCI. The first objective in this post hoc analysis was to compare the incidences of periprocedural bleeding in patients undergoing PCI between access routes. The second aim was to compare the incidence of bleeding between the 2 treatment groups.

Combined analyses of PRASFIT-ACS and PRASFIT-Elective were considered unfeasible because of the differences in the design of each study (eg, timing of the LD, and the option to

routes: the femoral and radial arteries.2–8 These studies have consistently shown better outcomes with the radial access route than with the femoral access route.2–8 P2Y12 inhibition is a cornerstone in the management of peri-PCI thrombotic events, and it is critically important to evaluate the effect of femoral or radial puncture on bleeding associated with prasugrel and clopidogrel. This question was not answered in the phase 3 TRITON-TIMI 38 study owing to low use of the radial access route.9 The Japanese PCI setting is unique because of the higher rate of use of the radial approach compared with the rest of the world. This is the first analysis to compare bleeding with standard and more intense regimens of P2Y12 inhibition via 2 access routes in Japan.

In this post hoc analysis, we pooled the results of the PRASugrel compared with clopidogrel For Japanese patIenTs

Table 2. Baseline PRASFIT-Elective Patient Characteristics vs. Artery Access Route

Variables Femoral Radial Brachial Multiple* P-value

n 135 508 60 8

Female 40 (29.6) 128 (25.2) 19 (31.7) 4 (50.0) 0.297†

Mean age (years) 68.1±10.1 67.4±8.8　　 66.1±10.3 67.6±8.7 0.465†

Age ≥75 years 48 (35.6) 108 (21.3) 14 (23.3) 2 (25.0) <0.001‡　Height (cm) 160.7±9.0　　　　 161.2±8.5　　　　 160.1±9.6　　　　 157.6±8.4　　 0.548†

BW (kg) 64.0±10.8 64.5±10.9 62.7±13.0 51.7±9.0 0.670†

BW ≤50 kg 16 (11.9) 37 (7.3) 10 (16.7) 4 (50.0) 0.086‡

BW ≤60 kg 43 (31.9) 182 (35.8) 26 (43.3) 7 (87.5) 0.389‡

BMI (kg/m2) 24.7±3.3　　 24.7±3.2　　 24.2±3.2　　 20.9±3.7 0.965†

Primary disease

SAP 100 (74.1) 383 (75.4) 53 (88.3) 4 (50.0) 0.753‡

Prior MI 8 (5.9) 26 (5.1) 1 (1.7) 1 (12.5) 0.709‡

Unstable angina 11 (8.1) 46 (9.1) 2 (3.3) 3 (37.5) 0.742‡

Asymptomatic myocardial ischemia 16 (11.9) 53 (10.4) 4 (6.7) 0 0.636‡

LD followed by PCI 93 (68.9) 368 (72.4) 44 (73.3) 6 (75.0) 0.415‡

Multiple treated lesions at initial PCI 72 (53.3) 185 (36.4) 15 (25.0) 3 (37.5) <0.001‡　Complications

Diabetes 45 (33.3) 54 (40.0) 22 (36.7) 3 (37.5) 0.160‡

Hyperlipidemia 114 (84.4) 414 (81.5) 47 (78.3) 6 (75.0) 0.427‡

Hypertension 109 (80.7) 418 (82.3) 44 (73.3) 4 (50.0) 0.679‡

Asymptomatic cerebral infarction 8 (5.9) 34 (6.7) 5 (8.3) 1 (12.5) 0.749‡

Concomitant drugs used at the time of starting treatment

PPI 70 (51.9) 240 (47.2) 37 (61.7) 4 (50.0) 0.341‡

Ca antagonist 63 (46.7) 247 (48.6) 26 (43.3) 3 (37.5) 0.686‡

Statin 109 (80.7) 341 (67.1) 33 (55.0) 4 (50.0) 0.002‡

ACE inhibitor 22 (16.3) 55 (10.8) 3 (5.0) 1 (12.5) 0.082‡

ARB 65 (48.1) 238 (46.9) 25 (41.7) 2 (25.0) 0.788‡

β-blocker 55 (40.7) 159 (31.3) 13 (21.7) 2 (25.0) 0.039‡

Insulin 12 (8.9) 39 (7.7) 1 (1.7) 0 0.643‡

CrCl (ml/min)

>60 86 (63.7) 321 (63.2) 37 (61.7) 6 (75.0) 0.802‡

30–60 27 (20.0) 91 (17.9) 8 (13.3) 1 (12.5)

<30 0 1 (0.2) 0 1 (12.5)

Missing 22 (16.3) 95 (18.7) 15 (25.0) 0

CYP2C19 phenotype n=77 n=382 n=50 n=6

EM 27 (35.1) 116 (30.4) 19 (38.0) 4 (66.7) <0.001‡　 IM 36 (46.8) 190 (49.7) 24 (48.0) 0

PM 14 (18.2) 76 (19.9) 7 (14.0) 2 (33.3)

Data given as mean ± SD or n (%). †t-test (femoral vs. radial); ‡χ2 test (femoral vs. radial). *Multiple access routes: femoral+radial, n=5; radial+brachial, n=3. MI, myocardial infarction; SAP, stable angina pectoris. Other abbreviations as in Table 1.



platelet aggregation, drugs that increase the risk of bleeding by inhibiting platelet aggregation or coagulation, thrombolytic drugs, and acidic non-steroidal anti-inflammatory drugs. The access site (femoral or radial) to be used was at the investiga-tor’s discretion.

The primary endpoint of both studies was the incidence of MACE at 24 weeks of treatment. MACE was defined as a composite of cardiovascular death, non-fatal myocardial infarction, and non-fatal ischemic stroke. Secondary endpoints included individual types of cardiovascular events and bleed-ing-related adverse events.

PRASFIT-ACSPRASFIT-ACS was a 24–48-week study in which 1,363 Japanese patients with ACS undergoing PCI were randomized and treated with either prasugrel (20/3.75 mg, n=685) or clop-idogrel (300/75 mg, n=678) both in combination with aspirin (first dose, 81–330 mg; subsequent doses, 81–100 mg).10 The LD was administered before PCI or up to 1 h after leaving the cardiac catheter laboratory in urgent cases. The MD was admin-istered once daily after breakfast, starting on the day after the LD. Men and women scheduled for coronary artery stenting were eligible if they satisfied the following inclusion criteria: age ≥20 years, presence of chest discomfort or ischemic

skip the LD in PRASFIT-Elective) and types of patients enrolled (ACS patients in PRASFIT-ACS and elective treat-ment of coronary artery disease [CAD] in PRASFIT-Elective). Therefore, we felt that combined analyses might mask differ-ences in bleeding complications between PCI performed in elective and ACS settings.

MethodsThe designs of the PRASFIT-ACS and PRASFIT-Elective studies are reported in more detail elsewhere.10,11 Both trials were randomized and performed in a double-blind manner. Both trials were approved by ethics committees at each par-ticipating institution and were conducted in accordance with national and international regulations. Patients in both stud-ies provided written informed consent to participate in the studies. Both studies were registered with the Japan Phar-maceutical Information Center (identifiers: PRASFIT-ACS, JapicCTI-101339; PRASFIT-Elective, JapicCTI-111550). Daiichi Sankyo funded both trials and maintained the study databases used in this post hoc analysis.

In both studies, patients were prohibited from using drugs deemed likely to influence study outcomes, including drugs that reduce the risk of cardiovascular events by inhibiting

Figure 1. Incidences of major or minor periprocedural bleeding according to the access route for PCI in PRASFIT-ACS. (A–C) Bleeding events according to femoral or radial artery access routes. (D–F) Bleeding events according to access route in 39 patients with multiple access routes (femoral+radial+brachial, n=1; femoral+radial, n=29; femoral+brachial, n=6; radial+brachial, n=3; femoral, n=36; radial, n=33). (A,D) Major or minor bleeding not related to CABG. (B,E) Major or minor bleeding as a compli-cation of PCI. (C,F) Major or minor bleeding at the puncture site. The number of patients with an event and the evaluable number are presented below each bar. *Fisher’s exact test.


1932 SAITO S et al.

Definitions of Periprocedural BleedingPeriprocedural bleeding events were defined as bleeding events not related to the coronary artery bypass graft (CABG), if the bleeding event occurred up to 3 days after the LD (PRASFIT-ACS) or up to 3 days after PCI (PRASFIT-Elective). The present analyses focused on a combination of major and minor bleeding using the TIMI definitions. Major bleeding was defined as intracranial or clinically significant bleeding with a decrease in hemoglobin ≥5 g/dl. Minor bleeding was defined as clini-cally significant bleeding with a decrease in hemoglobin 3–<5 g/dl. Bleeding events were evaluated by Safety Evalua-tion Committees in both studies.

Statistical AnalysisAll analyses were conducted in a post hoc manner in each study. For patient baseline characteristics and the incidences of major or minor periprocedural bleeding according to access route, we pooled the 2 treatment groups and compared the 2 access routes in patients who underwent PCI for first revascu-larization. These analyses were done using the safety analysis set, which consisted of all patients who took at least one dose of study drugs and did not have a serious violation of the study protocol. Baseline characteristics are presented descriptively as n (%) or mean ± SD in both studies. t-tests were used to

symptoms lasting ≥10 min within 72 h before randomization; ST-segment deviation ≥1 mm, or T-wave inversion ≥3 mm, or elevated cardiac biomarkers for necrosis.10 The major exclu-sion criteria are described in the original report.

PRASFIT-ElectivePRASFIT-Elective was a 24–48-week study in which 742 patients undergoing elective PCI for the treatment of CAD were randomized and treated with either prasugrel (20/3.75 mg) or clopidogrel (300/75 mg).11 The LD was administered 6–96 h before PCI, and the MD was administered once daily after breakfast. The investigator could also decide whether to administer MD without the LD, in which case the MD was administered for 14–21 days before PCI. Randomization occurred after the physician had decided whether to admin-ister the LD. All patients took aspirin (81–100 mg/day) start-ing at the time of the first dose of the study drug. Men and women were eligible for PRASFIT-Elective if they satisfied the following inclusion criteria: age ≥20 years; scheduled for elective PCI to treat CAD (eg, stable angina or prior myocar-dial infarction with stenosis confirmed on coronary computed tomography). The major exclusion criteria are described in the original report.

Figure 2. Incidences of major or minor bleeding according to access route in PRASFIT-Elective. (A–C) Bleeding events accord-ing to femoral or radial artery access routes. (D–F) Bleeding events according to access route in eight patients with multiple access routes (femoral+radial, n=5; radial+brachial, n=3; femoral, n=5; radial, n=8). (A,D) Non-CABG-related major or minor bleeding. (B,E) Major or minor bleeding as a complication of PCI. (C,F) Major or minor bleeding at the puncture site. The number of patients with an event and the evaluable number are presented below each bar. *Fisher’s exact test.



was used in nearly all patients. By contrast, in PRASFIT-Elective, the radial route was used in the majority of patients (n=508, 68.5%); the femoral route was used in 135 patients (18.2%). The brachial route was used in 35 patients in PRASFIT-ACS and in 60 patients in PRASFIT-Elective (75 patients in PRASFIT-ACS and 31 in PRASFIT-Elective under-went no or a different procedure). Multiple access routes were used in 39 patients in PRASFIT-ACS and in 8 patients in PRASFIT-Elective. In PRASFIT-ACS, there were significant differences between the femoral and radial groups regarding body weight, body mass index, time from onset of ST-seg-ment elevation myocardial infarction/chest discomfort/isch-emic symptoms to the start of PCI or administration of the LD in PRASFIT-ACS (Table 1). In PRASFIT-Elective, there were significant differences between the femoral and radial groups in the proportion of elderly patients, patients with ≥2 lesions, patients on concomitant statin therapy, and patients on con-comitant β-blocker therapy (Table 2).

compare continuous variables and chi-squared tests were used to compare categorical variables at baseline between the fem-oral and radial access routes. Univariate and multivariate Cox regression models were used to identify possible risk factors for periprocedural bleeding. The multivariate model was per-formed in a stepwise manner using all the variables included in univariate analysis. The variables included in the models were selected from the factors that were considered likely to affect bleeding. All statistical analyses were done using SAS 9.2 (SAS Institute, Cary, NC, USA). The corresponding author had full access to the study data and results.

ResultsVascular Access Route and Patient CharacteristicsThe characteristics of patients according to the access route are summarized in Tables 1,2 for PRASFIT-ACS and PRASFIT-Elective, respectively. In PRASFIT-ACS, the fem-oral (n=683, 50.1%) or the radial (n=531, 39.0%) access route

Table 3. Risk Factors for Periprocedural Bleeding in PRASFIT-ACS

Major/minor bleeding events HR 95% CI P-value

Univariate

Sex

Male (vs. female) 0.275 0.149–0.507 <0.0001　 BW

≤50 kg (vs. >50 kg) 4.091 2.145–7.802 <0.0001　 Age

≥75 years (vs. <75 years) 3.259 1.767–6.013 0.0002

eGFR 0.8325

Moderately decreased (vs. normal or mild) 1.388 0.674–2.857

Severely decreased or ESRF (vs. normal or mild) 0.000 0.000

Unknown (vs. normal or mild) 0.921 0.280–3.029

Hypertension

Yes (vs. no) 1.04 0.521–2.075 0.9110

Disease type 0.2290

STEMI (vs. UA or NSTEMI) 1.745 0.924–3.294

Other (vs. UA or NSTEMI) 0.000 0.000

Timing of the LD 0.2331

>6 h before first PCI (vs. after first PCI) 1.405 0.373–5.298

<6 h before first PCI (vs. after first PCI) 2.147 0.982–4.696

Between first PCI (vs. after first PCI) 0.471 0.059–3.768

No first PCI (vs. after first PCI) 0.000 0.000

Access route

Femoral artery Yes (vs. No) 3.739 1.727–8.094 0.0008

Radial artery Yes (vs. No) 0.288 0.128–0.650 0.0027

Treatment

Prasugrel (vs. clopidogrel) 1.408 0.756–2.621 0.2805

Multivariate

Sex

Male (vs. female) 0.418 0.208–0.843 0.0148

Age

≥75 years (vs. <75 years) 2.205 1.138–4.273 0.0192

Access route


Treatment

Prasugrel (vs. clopidogrel) 1.308 0.701–2.439 0.3985

CI, confidence interval; eGFR, estimated glomerular filtration rate; ESRF, end-stage renal failure; HR, hazard ratio; UA, unstable angina. Other abbreviations as in Table 1.


1934 SAITO S et al.

events. In the multivariate model, the risk of bleeding was lower with the radial artery access route compared with other arterial routes. We also conducted similar analyses in PRASFIT-Elective, and the results are summarized in Table 4. In that study, body weight (≤50 kg) and access route (femoral artery) were independent risk factors for periprocedural bleed-ing defined as major or minor bleeding events not related to CABG.

Rates of Periprocedural Bleeding vs. Treatment and Access RouteFigure 3 compares the rates of bleeding events between the prasugrel and clopidogrel groups according to the arterial access route used. In PRASFIT-ACS (Figure 3A), the inci-dence of minor bleeding was slightly but non-significantly greater in the prasugrel group than in the clopidogrel group (11/366, 3.0% vs. 4/353, 1.1%) in patients whose PCI was conducted via the femoral access route. Among those who underwent PCI via the radial access route, only 1.1% of patients (3/285) in the prasugrel group experienced minor bleeding compared with 0% for minor bleeding in the clopi-dogrel group and for major bleeding in both groups. In PRASFIT-Elective, minor bleeding occurred in only 2.7% of patients (2/75) treated with prasugrel and in 1.5% of patients (1/65) treated with clopidogrel following PCI via a femoral access route (Figure 3B). There were no major bleeding

Incidence of Periprocedural BleedingFigures 1,2 compare the rates of bleeding not related to CABG, bleeding as a complication of PCI, and bleeding related to the puncture site between the femoral and radial groups in PRASFIT-ACS (Figure 1) and in PRASFIT-Elective (Figure 2). In both studies, the rates of all 3 bleeding categories were consistently lower using a radial access route than with a femoral access route. This difference was also apparent when we considered patients who required multiple access routes.

Risk Factors for Periprocedural BleedingThe results of univariate and multivariate analyses of risk fac-tors for periprocedural bleeding are presented in Table 3 for PRASFIT-ACS and in Table 4 for PRASFIT-Elective. In PRASFIT-ACS, the risk of non-CABG-related major/minor bleeding events was significantly different among the sex, body weight, age, and access route categories on univariate analysis. In the multivariate model, which also included treat-ment received, the variables sex, age group, and access route were independently associated with the risk of major/minor bleeding. In univariate and multivariate models, the risk of bleeding was significantly greater for the femoral artery route compared with other arterial routes. In the analyses of the risk of major, minor, or clinically significant bleeding events, sex and access route were independent risk factors for these

Table 4. Risk Factors for Periprocedural Bleeding in PRASFIT-Elective

Major/minor bleeding events HR 95% CI P-value

Univariate

Sex

Male (vs. female) 0.368 0.052–2.613 0.3175

BW

≤50 kg (vs. >50 kg) 9.597 1.352–68.130 0.0237

Age

≥75 years (vs. <75 years) 9.384 0.976–90.211 0.0525

eGFR 0.9995

Moderate decrease (vs. normal or mild) 1.158 0.12–11.129

Severe decrease or ESRF (vs. normal or mild) 0.000

Unknown (vs. normal or mild) 0.000

Correct eGFR 1.0000

Moderate decreased (vs. normal or mild) >100 0.000

Severe decreased or ESRF (vs. normal or mild) 1.000 0.000

Unknown (vs. normal or mild) 1.000 0.000

Hypertension

Yes (vs. no) >100 0.9944

Stable angina

Yes (vs. no) >100 0.9939

Initial treatment

LD (vs. MD) 1.176 0.122–11.310 0.8881

Access route

Femoral artery Yes (vs. No) 12.32 1.282–>100 0.0296

Radial artery Yes (vs. No) 0.125 0.013–1.205 0.0721

Multivariate

BW

≤50 kg (vs. >50 kg) 7.261 1.001–52.675 0.0499

Access route


MD, maintenance dose. Other abbreviations as in Tables 1,3.



dence of periprocedural bleeding was similar between the prasugrel and clopidogrel groups, except for a slightly higher incidence in prasugrel-treated patients whose PCI was con-ducted via the femoral access route. In both studies, the choice of access route was at the discretion of the investigator. There-fore, it is interesting to note that the investigators were more likely to use the radial access route in elective settings, whereas PCI was performed via the radial and femoral routes in similar proportions of ACS patients.

One factor that might influence the choice of vascular access route is the time between the onset of symptoms and PCI. As shown in Table 1, a radial access route was more

events in either group for both access routes, and no minor events in either group when using the radial access routes.

DiscussionThere were 3 major findings of the post hoc analyses con-ducted here. First, the radial access route was more frequently used in PRASFIT-Elective than in PRASFIT-ACS. Second, independent risk factors for non-CABG-related major/minor periprocedural bleeding included body weight ≤50 kg and use of the femoral artery access route in both studies, together with female sex (vs. male) in PRASFIT-ACS. Third, the inci-

Figure 3. Incidence of major or minor bleeding at the puncture site according to access route and allocated treatment (prasugrel or clopidogrel) in (A,B) PRASFIT-ACS and (C,D) PRASFIT-Elective according to the (A,C) femoral or (B,D) radial artery access routes. The values in brackets on each bar represent the numbers of patients with an event. *Fisher’s exact test.


1936 SAITO S et al.

of other access routes and periprocedural bleeding. Prospec-tive, randomized trials may be necessary to confirm the pres-ent findings and eliminate these sources of bias.

ConclusionsThe main finding of these post hoc analyses of 2 large Japanese studies was that the femoral artery and radial artery access routes were used in similar proportions of ACS patients, whereas the radial artery route was the predominant route in elective settings. We also found that periprocedural bleeding was less frequently associated with the radial artery access route, and that risk factors for periprocedural bleeding included femoral artery access in both trials, age and sex in PRASFIT-ACS, and body weight in PRASFIT-Elective. Prasugrel was associated with a slightly greater incidence of minor bleeding for PCI via the femoral artery in both settings, whereas there were no such differences in the rates of major bleeding in procedures using the femoral artery, or in the rates of major/minor bleeding in procedures using the radial artery, associ-ated with prasugrel.

AcknowledgmentsThe authors thank Kimihiko Tanizawa, of Daiichi Sankyo, for assistance with study design, data acquisition and interpretation, and manuscript drafting and revision; and Nicholas D. Smith, PhD, for providing medical writing support. The authors also acknowledge the support of the institu-tions and investigators involved in the PRASFIT-ACS and PRASFIT-Elective studies.

DisclosuresS.S. is a medical advisor for Terumo and has received honoraria from Daiichi Sankyo, Abbot Vascular Japan, Boston Scientific Japan, and Medtronic.

T.I. has received honoraria and clinical research funding from Daiichi Sankyo, Otsuka Pharmaceutical, and Sanofi, and honoraria from AstraZeneca.

T.K. has received honoraria, clinical research funding, and other research funding from Daiichi Sankyo and Sanofi.

H.O. has received honoraria from Daiichi Sankyo, AstraZeneca, Bayer Yakuhin, Boehringer Ingelheim Japan, Bristol-Myers Squibb, Dainippon Sumitomo Pharma, Eisai, Kyowa Hakko Kirin, Mitsubishi Tanabe Pharma, MSD, Pfizer Japan, Sanofi, Takeda Pharmaceutical, and Teijin Pharma, clinical research funding from Daiichi Sankyo, and other research funding from Astellas Pharma, AstraZeneca, Boehringer Ingelheim Japan, Bristol-Myers Squibb, Chugai Pharmaceutical, Daiichi Sankyo, Dainippon Sumitomo Pharma, Eisai, Kowa, Mitsubishi Tanabe Pharma, MSD, Novartis Pharma, Otsuka Pharmaceutical, Pfizer Japan, Sanofi, Shionogi, and Takeda Pharmaceutical.

H.Y. has no conflicts of interest to declare.M. Nishikawa has received honoraria and clinical research funding

from Daiichi Sankyo and Otsuka Pharmaceutical.S.M. has received honoraria and clinical research funding from Daiichi

Sankyo and MSD.Y.T. is an employee of Daiichi Sankyo and performed the statistical

analyses.M. Nakamura has received honoraria from Daiichi Sankyo, AstraZeneca,

and Sanofi.

Financial SupportThe PRASFIT-ACS and PRASFIT-Elective studies were sponsored by Daiichi Sankyo (Tokyo, Japan).

References 1. Rao SV, Dai D, Subherwal S, Weintraub WS, Brindis RS, Messenger

JC, et al. Association between periprocedural bleeding and long-term outcomes following percutaneous coronary intervention in older patients. JACC Cardiovasc Interv 2012; 5: 958 – 965.

2. Jolly SS, Yusuf S, Cairns J, Niemela K, Xavier D, Widimsky P, et al. Radial versus femoral access for coronary angiography and inter-vention in patients with acute coronary syndromes (RIVAL): A

likely in patients with a longer time from onset of symptoms to PCI in ACS patients, possibly because the cardiologist had a longer time to plan the procedure and evaluate the suitability of radial access. The femoral route was more likely in patients with a shorter time from the onset of ACS symptoms to PCI or the LD (Table 1). In PRASFIT-Elective, the radial access route was more frequently used than the femoral access route (Table 2). This may be related to the amount of time available to plan the procedure. Some cardiologists may have preferred a radial access route based on the results of prior studies show-ing that coronary intervention via the radial access route may be associated with a lower incidence of MACE and bleeding compared with the femoral access route.7,8 The rate of major/minor periprocedural bleeding was somewhat greater in the PRASFIT-ACS study than in the PRASFIT-Elective study. Factors that might have contributed to these differences include the more frequent use of heparin and the more severe clinical condition of patients in PRASFIT-ACS.

The incidence of major/minor bleeding was consistently greater following PCI with a femoral approach than in PCI with a radial approach. This was also apparent in multivariate models, where the approach used was an independent predic-tor of bleeding, with the femoral approach being associated with a higher incidence of major/minor bleeding and the radial approach, with a lower incidence of bleeding in both PRASFIT-ACS (femoral access, hazard ratio [HR] 3.739; radial access, HR 0.288) and PRASFIT-Elective (femoral access, HR 12.32; radial access, 0.125).

The lower incidence of bleeding with the radial access route compared with the femoral access route in PRASFIT-ACS and PRASFIT-Elective is generally consistent with those of recent studies of non-Japanese patients,2–4,6,8 including a meta-analysis of 9 studies.7 Most of these studies were performed in ACS patients, and they consistently showed that the incidence of bleeding and other outcomes was greater if PCI was per-formed using a femoral access route, similar to the present post hoc analysis of PRASFIT-ACS.

The third main finding of this study was that the rate of bleeding was generally similar between prasugrel and clopi-dogrel. The only appreciable difference was a numerically higher incidence of minor periprocedural bleeding in the pra-sugrel group than in the clopidogrel group for PCI performed using the femoral access route in PRASFIT-ACS and PRASFIT-Elective. The incidence of minor bleeding was low in both studies. Based on the results of these studies, clinicians should be aware that prasugrel-treated patients may have a slightly higher risk of periprocedural bleeding than clopido-grel-treated patients. Future analyses may be needed to assess whether this translates into a higher risk of rehospitalization for bleeding during long-term treatment, MACE, or other adverse outcomes.1

Some limitations of the present study need to be discussed. First, the analyses were conducted in a post hoc manner and the studies were not specifically powered to detect differences in the rates of periprocedural bleeding between the access routes or treatment groups. In addition, the access route was not randomly allocated, which may introduce some bias. It is possible that the clinicians selected the access route after con-sidering patient factors likely to influence patient outcome, including ease of access in individual patients and the time available to plan the optimal access route. It is also possible that some clinicians, especially those treating ACS patients, were more familiar with the femoral access route. Finally, because other access routes were infrequently used, it was not possible to examine the possible associations between the use



103: 796 – 800. 7. Mamas MA, Ratib K, Routledge H, Fath-Ordoubadi F, Neyses L,

Louvard Y, et al. Influence of access site selection on PCI-related adverse events in patients with STEMI: Meta-analysis of randomised controlled trials. Heart 2012; 98: 303 – 311.

8. Lee MS, Wolfe M, Stone GW. Transradial versus transfemoral per-cutaneous coronary intervention in acute coronary syndromes: Re-evaluation of the current body of evidence. JACC Cardiovasc Interv 2013; 6: 1149 – 1152.

9. Wiviott SD, Braunwald E, McCabe CH, Montalescot G, Ruzyllo W, Gottlieb S, et al; TRITON-TIMI-38 investigators. Prasugrel versus clopidogrel in patients with acute coronary syndromes. N Engl J Med 2007; 357: 2001 – 2015.

10. Saito S, Isshiki T, Kimura T, Ogawa H, Yokoi H, Nanto S, et al. Efficacy and safety of adjusted-dose prasugrel compared with clopi-dogrel in Japanese patients with acute coronary syndrome: The PRASFIT-ACS study. Circ J 2014; 78: 1684 – 1692.

11. Isshiki T, Kimura K, Ogawa H, Yokoi H, Nanto S, Takayama M, et al. Prasugrel, a third-generation P2Y12 receptor antagonist, in patients with coronary artery diseases undergoing elective PCI: A Phase III, randomized, double-blind study. Circ J 2014; 78: 2926 – 2934.

randomised, parallel group, multicentre trial. Lancet 2011; 377: 1409 – 1420.

3. Mehta SR, Jolly SS, Cairns J, Niemela K, Rao SV, Cheema AN, et al. Effects of radial versus femoral artery access in patients with acute coronary syndromes with or without ST-segment elevation. J Am Coll Cardiol 2012; 60: 2490 – 2499.

4. Romagnoli E, Biondi-Zoccai G, Sciahbasi A, Politi L, Rigattieri S, Pendenza G, et al. Radial versus femoral randomized investigation in ST-segment elevation acute coronary syndrome: The RIFLE-STEACS (Radial Versus Femoral Randomized Investigation in ST-Elevation Acute Coronary Syndrome) study. J Am Coll Cardiol 2012; 60: 2481 – 2489.

5. Bernat I, Horak D, Stasek J, Mates M, Pesek J, Ostadal P, et al. ST-segment elevation myocardial infarction treated by radial or femoral approach in a multicenter randomized clinical trial: The STEMI-RADIAL trial. J Am Coll Cardiol 2014; 63: 964 – 972.

6. Sciahbasi A, Pristipino C, Ambrosio G, Sperduti I, Scabbia EV, Greco C, et al. Arterial access-site-related outcomes of patients undergoing invasive coronary procedures for acute coronary syn-dromes (from the ComPaRison of Early Invasive and Conservative Treatment in Patients With Non-ST-ElevatiOn Acute Coronary Syn-dromes [PRESTO-ACS] Vascular Substudy). Am J Cardiol 2009;

impact of arterial access route on bleeding complications

Documents